During the previous 2 cycles of this Program Project on "Molecular Biomarkers for Environmental Carcinogens," the investigators have established close links with the Projects which have led to several publications. As part of the renewal of the Program Project, they will continue to collaborate with investigators of Projects in the Program Project for the design of studies using specimens collected as part of previous studies, for the analysis of data, and for the development of methods relevant to the issues of the research Projects. The design and analysis of the experiments and studies to be conducted under the auspices of the Program Project will benefit by the inclusion of expertise in biostatistics and epidemiology. The specific aims of the Biostatistics/Epidemiology Core are: (1) To establish methods that promote adherence to standard protocols with particular attention to data collection and management. The Biostatistics/Epidemiology Core will assist with the selection of appropriate samples for testing; monitor data entry; implement data editing procedures design a data base to expedite analysis and linkage of the different projects; and establish systems for data security and backup. (2) To collaborate with investigators in the Projects of the Program Project for the purpose of data analysis. Data from the projects will require the implementation of multivariate (e.g., exposure to aflatoxin, HBV status and p53 mutation on codon 249) regression methods for analyses of longitudinal data and for binary and continuous outcomes. Furthermore, they will use methods for the evaluation of treatments administered in clinical trials using changes of biomarkers as the primary outcome measure. (3) To develop new statistical methods appropriate for the data generated by the Projects which will allow for testing and estimating parameters relevant to the specific aims of the Program Project. First, they will investigate the use of sequential designs to minimize sample testing. Second, a salient feature of the data from the proposed Projects is that most biomarkers will be a mixture of discrete (% undetectable) and continuous (level among those detectable) components. Methods that treat the data as left censored, as opposed to assigning all undetectable observations a given value, are preferred and will be developed and implemented as part of this application.